Predictive Maintenance For People: The Endgame For The Internet Of Things (#IoT)

Tom Raftery

Predictive maintenance is one of the oldest and most tested uses cases for the Internet of Things (IoT). For years, we’ve been able to analyze incoming data from sensors embedded in machines and make decisions about whether or not maintenance activities should be executed.

Typical scenarios have focused on things like wind farms, oil rigs, and fleets of trains. And while there’s plenty of excitement and new developments in these areas, what’s grabbing a lot of attention today is how predictive maintenance can be applied to new scenarios.

For example, in an earlier blog, I talked about predictive maintenance for autonomous vehicles – how sensors can send out data on the status of parts and components, allowing manufacturers to analyze this data to predict part failure and, thus, avoid breakdowns.

Yet, even this scenario keeps us in the realm of machines – because, as sophisticated as it may be, an autonomous vehicle is still a machine. But what if we could take the same general idea of predictive maintenance for machines and apply it to our bodies? Call it preventative maintenance for people – or just predictive healthcare. The reality is that in many ways, we’re already there.

Understanding in context

One of the advantages of predictive maintenance for machines is that incoming data about what’s going on in the moment can be analyzed in the context of historical data about the same machine. Let’s say an HVAC machine on the top of a hotel in Seville – where I live – sends out a high-temperature alert.

In and of itself, this may be cause for concern. But when you realize that the machine sends out the same alert every month at the same time – well, maybe it’s not so concerning. Maybe the HVAC unit runs continuously for eight hours on the first Monday of every month to help cool a large conference room for the packed monthly meeting of the Seville Dog Walker’s Association.

Or maybe there’s another reason. The point is that in such a scenario, the high temperature alert is understandable and predictable in context – and thus of little concern. It would be nice if we had something similar for healthcare.

More than a snapshot

On a typical trip to the doctor, you sit in the waiting room for 10-15 minutes with other people, some of whom are likely sick. When you finally see the doctor, you’re thinking of the next appointment you have across town in 30 minutes, so your anxiety levels go up.

My sister Mary was recently diagnosed with high blood pressure because her blood pressure measured 150/89 in the doctor’s clinic. The doctor advised her to get a connected blood pressure cuff and take regular measurements. When she did, it turned out her blood pressure was 108/75 – completely normal. She was suffering from what doctors call White Coat Syndrome.

Her high blood pressure reading was understandable in the context of her being in a doctor’s office, much like the HVAC data was skewed by a temporary situation. Wouldn’t it be great if the doctor had more than a snapshot of (often misleading) blood pressure data to work from? Wouldn’t a whole bunch of relevant historical data be better?

With the smartwatch on my wrist, I can now share three years of data with my doctor. Now she can see things in context and treat me more effectively. I think it’s only a matter time before her office can take my sensor data in automatically, over the cloud. This will make my yearly checkup more productive. Instead of figuring out what the problem is (if there is one, hopefully not), we’ll be able to focus on what to do about it.

A business network for health

As with so many things IoT, this is only the beginning. But let’s step back for a moment and look at the notion of an asset intelligence network (AIN).

Think of it as a business network application. All of the data (metadata, specifications, bills of materials, etc.) that goes into the creation of a device (compressor, coffee machine, car, etc.) can be stored in a central location.

When connected to the asset intelligence network, the device can push out real-time data that describes its state at any given moment. When the device owner allows access to this data, the manufacturer can then analyze it in conjunction with other data from other devices – making product improvements that can then be pushed out by way of the same asset intelligence network.

In fact, nothing is stopping device owners from sharing their data with anyone they wish – such as a service vendor or an insurance company. If a device goes out of tolerance for some reason, the service vendor could receive a notification and schedule an appointment to service the device automatically. Or an insurance company could set rates according to actual device usage data.

Returning to the theme of health, what if we took this idea of an asset intelligence network and applied it to our own bodies? What if we had a “people’s intelligence network” where a device like my smartwatch publishes my health data into a trusted cloud application? When my device senses high blood sugar, for example, this data is analyzed, not only in the context of the unique moment mixed with my own personal health history, but also in the context of similar data from potentially millions of people.

Based on this much larger dataset, the network could then contact my service vendor – in this case, my doctor – and make an appointment if necessary. Yes, this would be convenient. But more importantly, it would move us away from making medical decisions based on poor data and the intuition of physicians, toward something often heralded but seldom achieved – real evidence based medicine.

Learn more about the SAP Asset Intelligence Network.

This article was originally posted on TomRaftery.com. Follow me on Twitter @TomRaftery


Tom Raftery

About Tom Raftery

Tom Raftery is Vice President and Global Evangelist for the Internet of Things at SAP. Previously, Tom worked as an independent analyst focusing on the Internet of Things, energy, and clean technology. Tom has a very strong background in social media, is the former co-founder of a software firm, and is co-founder and director of hyper energy-efficient data center Cork Internet eXchange. More recently, Tom worked as an industry analyst for RedMonk, leading the GreenMonk practice for seven years.

The Challenge Of Analytics Growth In The Public Sector

Shaily Kumar

Although the opportunities to apply analytics in the public sector are abundant, cultural and technical challenges must be overcome before government agencies can claim to be fully developed, enterprise-wide, analytically competitive organizations. Building an analytical culture where data is widely used to evaluate hypotheses is crucial for an analytically competitive organization.

Despite the successes that the public sector has seen in the past with analytics, data analysis is not integrated into most decision-making processes. This can partly be attributed to the enormous variety of tasks in many different fields that government organizations perform. In such varied environments, one-size-fits-all approach to cultural change is often ineffective, and customized approaches training, policies, and incentives are necessary. These possible solutions require time and effort.

In both the public and private sectors, a scarcity of analytical talent makes it difficult for an analytical culture to flourish. Furthermore, unlike the private sector, many components of the public sector cannot outsource analytical work to contracting companies due to security and privacy issues. While competing to hire talent in public sector is difficult, it is not impossible; disadvantages in compensation (compared to financial and other quantitative industries that recruit analytics professionals) can be balanced by focusing on the importance of the mission: meaningful projects around public services.

The public sector (for example, the national security agency) manages a diverse and fast-changing set of risks and challenges. Terrorism, climate change, inequalities in natural resources, cybercrime, and the decline of the critical national infrastructure are just some of the challenges the agencies are trying to overcome to ensure the safety and prosperity of their countries. Meeting these challenges while juggling budget cuts requires a cohesive preventative approach across agencies like defense, police, intelligence, and border agencies. To succeed, a step change is needed in the way public sector organizations manage, share, and exploit their information.

The London Fire Brigade (LFB) has adopted a unique approach to integrate external data with its own and use predictive analytics to optimize performance. Internal records of where fires have occurred in the past are now integrated with a range of external data – including census information, land use, and lifestyles – to produce future fire risk maps of London. The maps inform budget allocation and enable the LFB to target its limited firefighting and preventative resources to the buildings and people most at risk. Put simply, using predictive analytics literally avoids the need for expensive and high-risk firefighting across the public sector.

Analytics in the public sector must also be responsive to unequal stakeholder groups. It is necessary, but not sufficient, to design analysis to satisfy the primary customer (often a decision-maker higher up in the chain of command); the wishes of the Parliament and the public must be considered, and they generally do not have a consensus on their objectives. When analytical findings are observed to favor one policy alternative over another, stakeholders may also challenge the analysis process to protect their positions and beliefs.

Outside the federal government, state and local governments are also working on these challenges, with varying levels of progress. Many of these organizations need the critical mass of analytical professionals to integrate analysis into all decision-making, and only a few have enough analytical talent to make substantial progress. At the local level, some innovative approaches have started, such as the New York City Police Department’s CompStat program, a data-driven view of police action that has extended to several major cities and has been extended with additional service functions. In the future, the “laboratories of democracy” that form an integral part of the American governance system will continue to develop new ways of using and approaches to analysis.

Finally, measuring the impact of analytics in the public sector is difficult because government success statistics are far more complex than simpler measures from the private sector. For example, counterterrorism decisions aim to minimize the number of fatalities, injuries, and economic consequences of an attack, within the limits of cost and impact on legitimate trade. Reasonable people agree that all these goals are important, but they do not agree on the values of the trade-offs between the objectives, so it is impractical to construct a single utility function to inform these decisions.

Efforts at all levels of the public sector give hope that analytics will be better integrated into decision-making, despite the challenges described above. The growing number of experienced analysts, the increasing availability of analytical tools that are easy to use and provide quick, easy-to-understand results; the accumulation of data and the dissemination of success stories all point to better integration. With many public-sector agencies giving priority to the development of analytical capabilities, analytics in the public sector will continue to grow, but this growth will require a constant effort to meet the challenges posed along the way.

For more on this topic, see Analytics In The Public Sector.


Shaily Kumar

About Shaily Kumar

Shailendra has been on a quest to help organisations make money out of data and has generated an incremental value of over one billion dollars through analytics and cognitive processes. With a global experience of more than two decades, Shailendra has worked with a myriad of Corporations, Consulting Services and Software Companies in various industries like Retail, Telecommunications, Financial Services and Travel - to help them realise incremental value hidden in zettabytes of data. He has published multiple articles in international journals about Analytics and Cognitive Solutions; and recently published “Making Money out of Data” which showcases five business stories from various industries on how successful companies make millions of dollars in incremental value using analytics. Prior to joining SAP, Shailendra was Partner / Analytics & Cognitive Leader, Asia at IBM where he drove the cognitive business across Asia. Before joining IBM, he was the Managing Director and Analytics Lead at Accenture delivering value to its clients across Australia and New Zealand. Coming from the industry, Shailendra held key Executive positions driving analytics at Woolworths and Coles in the past. Please feel to connect on: Linkedin: http://linkedin.com/in/shaily Twitter: https://twitter.com/meisshaily

Monetizing And Optimizing Content Distribution With Machine Learning And Blockchain

Catherine Lynch

A generation ago, media conglomerates tightly controlled content production and distribution, deciding when, where, and how content was consumed. That’s all changed. Gone are the days of linear television channels and a single-television household. Today’s consumers decide when and where to consume content across multiple platforms.

With the average attention span of an adult hovering at eight seconds, down from 15 seconds in 2000, the media industry is fighting for increasingly smaller slivers of consumer attention. Media companies need a solution for monetizing content and delivering the right content to the right consumer at the right moment. Advanced analytics, machine learning, and blockchain are three disruptive technologies that can solve the twin problems of volume overload and content monetization.

How advanced analytics and machine learning solve the “paradox of choice”

For media companies, consumers drive demand. It’s all about what they want, when they want it, and which device they want it on.

“We went from a very analog-driven, subscriber numbers rated world to a world where it’s about engagement, and about data, and about consuming the content when you want,” says Richard Whittington, senior vice president, Media Industry Cloud Solutions, in the S.M.A.C. Talk Technology Podcast.

Of course, if consumers can’t find they content they want, they can’t consume it. In a world with nearly infinite choices, consumers are increasingly paralyzed by the “paradox of choice.” This theory states that there is a tipping point for choice, a point where more choices cease to provide an advantage and instead become a hindrance. It’s akin to the feeling of mindlessly scrolling Netflix looking for something to watch, but not finding anything. One-third of consumers say that they frequently cannot find anything to watch, according to a Cord Cutting Survey conducted for Rovi.

Media companies are experimenting with new machine learning algorithms to better understand consumer behavior, preference, and social cues. With machine learning is it easier to utilize metadata through intuitive, creative applications, rather than simply recommending a movie based on genre or actor preference. For example, machine learning enables language processing for a deeper understanding of content based on mood, emotion or intensity. Coupled with social signals, such as a conversation on Facebook around a new movie, machine-learning powered content recommendations could boost viewer engagement, satisfaction, and loyalty. Relevancy and timing are paramount: media companies that can provide consumers with a perfectly curated shortlist may outperform the companies offering an endless list of options that miss the mark.

New monetization pathways with blockchain and machine learning

Since consumers moved away from physical products like CDs or DVDs, media companies have struggled to monetize their content. According to Whittington, blockchain offers a new path forward, addressing problems associated with rights management, payment, and distribution.

“Blockchain gives media companies the ability to track content and create events when content is consumed,” says Whittington in the S.M.A.C. Talk Technology Podcast. “For example, if I send you a football match to view, this will trigger an event that indicates that you consumed the match. Money is then paid to whoever owns the rights to that match, rather than having to go through the traditional controlled, linear model. Blockchain has the ability to turn the whole business model upside down.”

In addition to using blockchain to monetize content distribution and consumption, Whittington says machine learning may also play an important role in content monetization.

“We’ve always heard about product placement in shows,” says Whittington in the S.M.A.C. Talk Technology Podcast. “But [product placement] has never been able to be measured to such an extent with heat mapping and knowing exactly what people looked at, and did they notice it, and how long did they look at it for, and what is the value of those impressions compared to other media avenues that they might have put those dollars into. I love the example, for instance, of how can you use machine learning to say, hey, on this episode of ‘Modern Family’ we had this many times that we showed XYZ’s product and we’re going to charge you for this. If you don’t see any value in this company A, company B might. We can actually create competition there.”

Next steps: Gaining the first-mover advantage

The media industry has undergone a transformation from a distribution model to a direct-to-consumer model and continues to evolve at a rapid pace. By 2020, Gartner predicts that artificial intelligence (AI) bots, rather than humans, will manage 85 percent of customer interactions. There will be over 82 million US millennial digital video customers. As media companies grapple with the challenge of getting the right content to the right consumer at the right time, companies that proactively invest in advanced analytics, machine learning and blockchain will gain a critical first-mover advantage. These companies will be best positioned to turn data into insights, monetizing the delivery of the right piece of content at the right moment to the right consumer.

To learn more about how digital transformation is disrupting content distribution and monetization in the media industry, listen to the S.M.A.C. Talk Technology Podcast with Richard Whittington.

Hear the full podcast episode here. For more insight on digital leaders, check out the SAP Center for Business Insight report, conducted in collaboration with Oxford Economics, “SAP Digital Transformation Executive Study: 4 Ways Leaders Set Themselves Apart.”


Catherine Lynch

About Catherine Lynch

Catherine Lynch is a Senior Director of Industry Cloud Marketing at SAP. She is a content marketing specialist with a particular focus on the professional services and media industries globally. Catherine has a wide international experience of working with enterprise application vendors in global roles, creating thought leadership and is a social media practitioner.

Hack the CIO

By Thomas Saueressig, Timo Elliott, Sam Yen, and Bennett Voyles

For nerds, the weeks right before finals are a Cinderella moment. Suddenly they’re stars. Pocket protectors are fashionable; people find their jokes a whole lot funnier; Dungeons & Dragons sounds cool.

Many CIOs are enjoying this kind of moment now, as companies everywhere face the business equivalent of a final exam for a vital class they have managed to mostly avoid so far: digital transformation.

But as always, there is a limit to nerdy magic. No matter how helpful CIOs try to be, their classmates still won’t pass if they don’t learn the material. With IT increasingly central to every business—from the customer experience to the offering to the business model itself—we all need to start thinking like CIOs.

Pass the digital transformation exam, and you probably have a bright future ahead. A recent SAP-Oxford Economics study of 3,100 organizations in a variety of industries across 17 countries found that the companies that have taken the lead in digital transformation earn higher profits and revenues and have more competitive differentiation than their peers. They also expect 23% more revenue growth from their digital initiatives over the next two years—an estimate 2.5 to 4 times larger than the average company’s.

But the market is grading on a steep curve: this same SAP-Oxford study found that only 3% have completed some degree of digital transformation across their organization. Other surveys also suggest that most companies won’t be graduating anytime soon: in one recent survey of 450 heads of digital transformation for enterprises in the United States, United Kingdom, France, and Germany by technology company Couchbase, 90% agreed that most digital projects fail to meet expectations and deliver only incremental improvements. Worse: over half (54%) believe that organizations that don’t succeed with their transformation project will fail or be absorbed by a savvier competitor within four years.

Companies that are making the grade understand that unlike earlier technical advances, digital transformation doesn’t just support the business, it’s the future of the business. That’s why 60% of digital leading companies have entrusted the leadership of their transformation to their CIO, and that’s why experts say businesspeople must do more than have a vague understanding of the technology. They must also master a way of thinking and looking at business challenges that is unfamiliar to most people outside the IT department.

In other words, if you don’t think like a CIO yet, now is a very good time to learn.

However, given that you probably don’t have a spare 15 years to learn what your CIO knows, we asked the experts what makes CIO thinking distinctive. Here are the top eight mind hacks.

1. Think in Systems

A lot of businesspeople are used to seeing their organization as a series of loosely joined silos. But in the world of digital business, everything is part of a larger system.

CIOs have known for a long time that smart processes win. Whether they were installing enterprise resource planning systems or working with the business to imagine the customer’s journey, they always had to think in holistic ways that crossed traditional departmental, functional, and operational boundaries.

Unlike other business leaders, CIOs spend their careers looking across systems. Why did our supply chain go down? How can we support this new business initiative beyond a single department or function? Now supported by end-to-end process methodologies such as design thinking, good CIOs have developed a way of looking at the company that can lead to radical simplifications that can reduce cost and improve performance at the same time.

They are also used to thinking beyond temporal boundaries. “This idea that the power of technology doubles every two years means that as you’re planning ahead you can’t think in terms of a linear process, you have to think in terms of huge jumps,” says Jay Ferro, CIO of TransPerfect, a New York–based global translation firm.

No wonder the SAP-Oxford transformation study found that one of the values transformational leaders shared was a tendency to look beyond silos and view the digital transformation as a company-wide initiative.

This will come in handy because in digital transformation, not only do business processes evolve but the company’s entire value proposition changes, says Jeanne Ross, principal research scientist at the Center for Information Systems Research at the Massachusetts Institute of Technology (MIT). “It either already has or it’s going to, because digital technologies make things possible that weren’t possible before,” she explains.

2. Work in Diverse Teams

When it comes to large projects, CIOs have always needed input from a diverse collection of businesspeople to be successful. The best have developed ways to convince and cajole reluctant participants to come to the table. They seek out technology enthusiasts in the business and those who are respected by their peers to help build passion and commitment among the halfhearted.

Digital transformation amps up the urgency for building diverse teams even further. “A small, focused group simply won’t have the same breadth of perspective as a team that includes a salesperson and a service person and a development person, as well as an IT person,” says Ross.

At Lenovo, the global technology giant, many of these cross-functional teams become so used to working together that it’s hard to tell where each member originally belonged: “You can’t tell who is business or IT; you can’t tell who is product, IT, or design,” says the company’s CIO, Arthur Hu.

One interesting corollary of this trend toward broader teamwork is that talent is a priority among digital leaders: they spend more on training their employees and partners than ordinary companies, as well as on hiring the people they need, according to the SAP-Oxford Economics survey. They’re also already being rewarded for their faith in their teams: 71% of leaders say that their successful digital transformation has made it easier for them to attract and retain talent, and 64% say that their employees are now more engaged than they were before the transformation.

3. Become a Consultant

Good CIOs have long needed to be internal consultants to the business. Ever since technology moved out of the glasshouse and onto employees’ desks, CIOs have not only needed a deep understanding of the goals of a given project but also to make sure that the project didn’t stray from those goals, even after the businesspeople who had ordered the project went back to their day jobs. “Businesspeople didn’t really need to get into the details of what IT was really doing,” recalls Ferro. “They just had a set of demands and said, ‘Hey, IT, go do that.’”

Now software has become so integral to the business that nobody can afford to walk away. Businesspeople must join the ranks of the IT consultants.

But that was then. Now software has become so integral to the business that nobody can afford to walk away. Businesspeople must join the ranks of the IT consultants. “If you’re building a house, you don’t just disappear for six months and come back and go, ‘Oh, it looks pretty good,’” says Ferro. “You’re on that work site constantly and all of a sudden you’re looking at something, going, ‘Well, that looked really good on the blueprint, not sure it makes sense in reality. Let’s move that over six feet.’ Or, ‘I don’t know if I like that anymore.’ It’s really not much different in application development or for IT or technical projects, where on paper it looked really good and three weeks in, in that second sprint, you’re going, ‘Oh, now that I look at it, that’s really stupid.’”

4. Learn Horizontal Leadership

CIOs have always needed the ability to educate and influence other leaders that they don’t directly control. For major IT projects to be successful, they need other leaders to contribute budget, time, and resources from multiple areas of the business.

It’s a kind of horizontal leadership that will become critical for businesspeople to acquire in digital transformation. “The leadership role becomes one much more of coaching others across the organization—encouraging people to be creative, making sure everybody knows how to use data well,” Ross says.

In this team-based environment, having all the answers becomes less important. “It used to be that the best business executives and leaders had the best answers. Today that is no longer the case,” observes Gary Cokins, a technology consultant who focuses on analytics-based performance management. “Increasingly, it’s the executives and leaders who ask the best questions. There is too much volatility and uncertainty for them to rely on their intuition or past experiences.”

Many experts expect this trend to continue as the confluence of automation and data keeps chipping away at the organizational pyramid. “Hierarchical, command-and-control leadership will become obsolete,” says Edward Hess, professor of business administration and Batten executive-in-residence at the Darden School of Business at the University of Virginia. “Flatter, distributive leadership via teams will become the dominant structure.”

5. Understand Process Design

When business processes were simpler, IT could analyze the process and improve it without input from the business. But today many processes are triggered on the fly by the customer, making a seamless customer experience more difficult to build without the benefit of a larger, multifunctional team. In a highly digitalized organization like Amazon, which releases thousands of new software programs each year, IT can no longer do it all.

While businesspeople aren’t expected to start coding, their involvement in process design is crucial. One of the techniques that many organizations have adopted to help IT and businesspeople visualize business processes together is design thinking (for more on design thinking techniques, see “A Cult of Creation“).

Customers aren’t the only ones who benefit from better processes. Among the 100 companies the SAP-Oxford Economics researchers have identified as digital leaders, two-thirds say that they are making their employees’ lives easier by eliminating process roadblocks that interfere with their ability to do their jobs. Ninety percent of leaders surveyed expect to see value from these projects in the next two years alone.

6. Learn to Keep Learning

The ability to learn and keep learning has been a part of IT from the start. Since the first mainframes in the 1950s, technologists have understood that they need to keep reinventing themselves and their skills to adapt to the changes around them.

Now that’s starting to become part of other job descriptions too. Many companies are investing in teaching their employees new digital skills. One South American auto products company, for example, has created a custom-education institute that trained 20,000 employees and partner-employees in 2016. In addition to training current staff, many leading digital companies are also hiring new employees and creating new roles, such as a chief robotics officer, to support their digital transformation efforts.

Nicolas van Zeebroeck, professor of information systems and digital business innovation at the Solvay Brussels School of Economics and Management at the Free University of Brussels, says that he expects the ability to learn quickly will remain crucial. “If I had to think of one critical skill,” he explains, “I would have to say it’s the ability to learn and keep learning—the ability to challenge the status quo and question what you take for granted.”

7. Fail Smarter

Traditionally, CIOs tended to be good at thinking through tests that would allow the company to experiment with new technology without risking the entire network.

This is another unfamiliar skill that smart managers are trying to pick up. “There’s a lot of trial and error in the best companies right now,” notes MIT’s Ross. But there’s a catch, she adds. “Most companies aren’t designed for trial and error—they’re trying to avoid an error,” she says.

To learn how to do it better, take your lead from IT, where many people have already learned to work in small, innovative teams that use agile development principles, advises Ross.

For example, business managers must learn how to think in terms of a minimum viable product: build a simple version of what you have in mind, test it, and if it works start building. You don’t build the whole thing at once anymore.… It’s really important to build things incrementally,” Ross says.

Flexibility and the ability to capitalize on accidental discoveries during experimentation are more important than having a concrete project plan, says Ross. At Spotify, the music service, and CarMax, the used-car retailer, change is driven not from the center but from small teams that have developed something new. “The thing you have to get comfortable with is not having the formalized plan that we would have traditionally relied on, because as soon as you insist on that, you limit your ability to keep learning,” Ross warns.

8. Understand the True Cost—and Speed—of Data

Gut instincts have never had much to do with being a CIO; now they should have less to do with being an ordinary manager as well, as data becomes more important.

As part of that calculation, businesspeople must have the ability to analyze the value of the data that they seek. “You’ll need to apply a pinch of knowledge salt to your data,” advises Solvay’s van Zeebroeck. “What really matters is the ability not just to tap into data but to see what is behind the data. Is it a fair representation? Is it impartial?”

Increasingly, businesspeople will need to do their analysis in real time, just as CIOs have always had to manage live systems and processes. Moving toward real-time reports and away from paper-based decisions increases accuracy and effectiveness—and leaves less time for long meetings and PowerPoint presentations (let us all rejoice).

Not Every CIO Is Ready

Of course, not all CIOs are ready for these changes. Just as high school has a lot of false positives—genius nerds who turn out to be merely nearsighted—so there are many CIOs who aren’t good role models for transformation.

Success as a CIO these days requires more than delivering near-perfect uptime, says Lenovo’s Hu. You need to be able to understand the business as well. Some CIOs simply don’t have all the business skills that are needed to succeed in the transformation. Others lack the internal clout: a 2016 KPMG study found that only 34% of CIOs report directly to the CEO.

This lack of a strategic perspective is holding back digital transformation at many organizations. They approach digital transformation as a cool, one-off project: we’re going to put this new mobile app in place and we’re done. But that’s not a systematic approach; it’s an island of innovation that doesn’t join up with the other islands of innovation. In the longer term, this kind of development creates more problems than it fixes.

Such organizations are not building in the capacity for change; they’re trying to get away with just doing it once rather than thinking about how they’re going to use digitalization as a means to constantly experiment and become a better company over the long term.

As a result, in some companies, the most interesting tech developments are happening despite IT, not because of it. “There’s an alarming digital divide within many companies. Marketers are developing nimble software to give customers an engaging, personalized experience, while IT departments remain focused on the legacy infrastructure. The front and back ends aren’t working together, resulting in appealing web sites and apps that don’t quite deliver,” writes George Colony, founder, chairman, and CEO of Forrester Research, in the MIT Sloan Management Review.

Thanks to cloud computing and easier development tools, many departments are developing on their own, without IT’s support. These days, anybody with a credit card can do it.

Traditionally, IT departments looked askance at these kinds of do-it-yourself shadow IT programs, but that’s changing. Ferro, for one, says that it’s better to look at those teams not as rogue groups but as people who are trying to help. “It’s less about ‘Hey, something’s escaped,’ and more about ‘No, we just actually grew our capacity and grew our ability to innovate,’” he explains.

“I don’t like the term ‘shadow IT,’” agrees Lenovo’s Hu. “I think it’s an artifact of a very traditional CIO team. If you think of it as shadow IT, you’re out of step with reality,” he says.

The reality today is that a company needs both a strong IT department and strong digital capacities outside its IT department. If the relationship is good, the CIO and IT become valuable allies in helping businesspeople add digital capabilities without disrupting or duplicating existing IT infrastructure.

If a company already has strong digital capacities, it should be able to move forward quickly, according to Ross. But many companies are still playing catch-up and aren’t even ready to begin transforming, as the SAP-Oxford Economics survey shows.

For enterprises where business and IT are unable to get their collective act together, Ross predicts that the next few years will be rough. “I think these companies ought to panic,” she says. D!


About the Authors

Thomas Saueressig is Chief Information Officer at SAP.

Timo Elliott is an Innovation Evangelist at SAP.

Sam Yen is Chief Design Officer at SAP and Managing Director of SAP Labs.

Bennett Voyles is a Berlin-based business writer.

Read more thought provoking articles in the latest issue of the Digitalist Magazine, Executive Quarterly.

Tags:

Survey: Four Ways Machine Learning Will Disrupt Your Business

Dan Wellers and Dirk Jendroska

We are entering the era of the machine learning enterprise, in which this subset of artificial intelligence (AI) capabilities will revolutionize operating models, shake up staffing methods, upend business models, and potentially alter the nature of competition itself. The adoption of machine learning capabilities will be limited only by an organization’s ability to change – but not every company will be willing or able to make such a radical shift.

Very soon, the difference between the haves and the have-nots of machine learning will become clear. “The disruption over the next three to five years will be massive,” says Cliff Justice, principal in KPMG’s Innovation and Enterprise Solutions team. Companies hanging onto their legacy processes will struggle to compete with machine learning enterprises able to compete with a fraction of the resources and entirely new value propositions.

For those seeking to be on the right side of the disruption, a new survey, conducted by SAP and the Economist Intelligence Unit (EIU), offers a closer look at organizations we’ve identified as the Fast Learners of machine learning: those that are already seeing benefits from their implementations.

Machine learning is unlike traditional programmed software. Machine learning software actually gets better – autonomously and continuously – at executing tasks and business processes. This creates opportunities for deeper insight, non-linear growth, and levels of innovation previously unseen.

Given that, it’s not surprising that machine learning has evolved from hype to have-to-have for the enterprise in seemingly record time. According to the SAP/EIU survey, more than two-thirds of respondents (68%) are already experimenting with it. What’s more, many of these organizations are seeing significantly improved performance across the breadth of their operations as a result, and some are aiming to remake their businesses on the back of these singular, new capabilities.

So, what makes machine learning so disruptive? Based on our analysis of the survey data and our own research, we see four primary reasons:

1. It’s probabilistic, not programmed

Machine learning uses sophisticated algorithms to enable computers to “learn” from large amounts of data and take action based on data analysis rather than being explicitly programmed to do something. Put simply, the machine can learn from experience; coded software does not. “It operates more like a human does in terms of how it formulates its conclusions,” says Justice.

That means that machine learning will provide more than just a one-time improvement in process and productivity; those improvements will continue over time, remaking business processes and potentially creating new business models along the way.

2. It creates exponential efficiency

When companies integrate machine learning into business processes, they not only increase efficiency, they are able to scale up without a corresponding increase in overhead. If you get 5,000 loan applications one month and 20,000 the next month, it’s not a problem, says Sudir Jha, head of product management and strategy for Infosys; the machines can handle it.

3. It frees up capital – financial and human

Because machine learning can be used to automate any repetitive task, it enables companies to redeploy resources to areas that make the organization more competitive, says Justice. It also frees up the employees within an organization to perform higher-value, more rewarding work. That leads to reduced turnover and higher employee satisfaction. And studies show that happier employees lead to higher customer satisfaction and better business results.

4. It creates new opportunities

AI and machine learning can offer richer insight, deeper knowledge, and predictions that would not be possible otherwise. Machine learning can enable not only new processes, but entirely new business models or value propositions for customers – “opportunities that would not be possible with just human intelligence,” says Justice. “AI impacts the business model in a much more disruptive way than cloud or any other disruption we’ve seen in our lifetimes.”

Machine learning systems alone, however, will not transform the enterprise. The singular opportunities enabled by these capabilities will only occur for companies that dedicate themselves to making machine learning part of a larger digital transformation strategy. The results of the SAP/EIU survey explain the makeup of the evolving machine learning enterprise. We’ve identified key traits important to the success of these machine-learning leaders that can serve as a template for others as well as an overview of the outcomes they’re already seeing from their efforts.

Learn more and download the full study here.  

 


Dan Wellers

About Dan Wellers

Dan Wellers is founder and leader of Digital Futures at SAP, a strategic insights and thought leadership discipline that explores how digital technologies drive exponential change in business and society.

Dirk Jendroska

About Dirk Jendroska

Dr. Dirk Jendroska is Head of Strategy and Operations Machine Learning at SAP. He supports the vision of SAP Leonardo Machine Learning to enable the intelligent enterprise by making enterprise applications intelligent. He leads a team working on machine learning strategy, marketing and communications.